A randomised controlled trial to assess the clinical effectiveness and cost-effectiveness of alternative treatments to Inhibit VEGF in Age-related choroidal Neovascularisation (IVAN)

(1)

VOLUME 19 ISSUE 78 OCTOBER 2015 ISSN 1366-5278

A randomised controlled trial to assess the clinical

effectiveness and cost-effectiveness of alternative

treatments to Inhibit VEGF in Age-related choroidal

Neovascularisation (IVAN)

Usha Chakravarthy, Simon P Harding, Chris A Rogers, Susan Downes,

Andrew J Lotery, Helen A Dakin, Lucy Culliford, Lauren J Scott,

Rachel L Nash, Jodi Taylor, Alyson Muldrew, Jayashree Sahni,

(2)

(3)

the clinical effectiveness and

cost-effectiveness of alternative

treatments to Inhibit VEGF in Age-related

choroidal Neovascularisation (IVAN)

Usha Chakravarthy,

1 *

_{Simon P Harding,}

2 Chris A Rogers,

3 _{Susan Downes,}

4 _{Andrew J Lotery,}

5 Helen A Dakin,

6 _{Lucy Culliford,}

3 _{Lauren J Scott,}

3 Rachel L Nash,

3 _{Jodi Taylor,}

3 _{Alyson Muldrew,}

1 Jayashree Sahni,

2 _{Sarah Wordsworth,}

6 _{James Raftery,}

7 Tunde Peto

8 _{and Barnaby C Reeves}

3 _{for the}

IVAN Investigators

†

1

_{Centre for Experimental Medicine, Institute of Clinical Science,}

Queen

’

s University Belfast, Belfast, UK

2

_{Department of Eye and Vision Science, Institute of Ageing and Chronic Disease,}

University of Liverpool, Liverpool, UK

3

_{Clinical Trials and Evaluation Unit, School of Clinical Sciences, University of}

Bristol, Bristol, UK

4

_{Oxford University Hospitals NHS Trust, Oxford, UK}

5

_{Clinical and Experimental Sciences, Faculty of Medicine, University of}

Southampton, Southampton, UK

6

_{Health Economic Research Centre, Nuffield Department of Population Health,}

University of Oxford, Oxford, UK

7

_{Wessex Institute, University of Southampton, Southampton, UK}

8

_{National Institute for Health Research (NIHR) Biomedical Research Centre at}

Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of

Ophthalmology, London, UK

*Corresponding author

(4)

(5)

Bayer, Neovista, Oraya, Allergan, and Bausch and Lomb, and her employing institution has received payments from Novartis, Bayer, Neovista, Oraya, Alcon and Pfizer. Chris A Rogers has received an

honorarium from Novartis for a lecture. The employing institutions of Susan Downes and Andrew J Lotery have received payments from Novartis. Susan Downes and Andrew J Lotery have received honoraria from Novartis for lectures. Andrew J Lotery has attended and been remunerated for attendance at advisory boards for Novartis and Bayer. Barnaby C Reeves has received a fee for teaching from Janssen-Cilag and is a member of the National Institute of Health Research (NIHR) Health Technology Assessment

commissioning board and the NIHR Systematic Reviews Programme Advisory Group. James Raftery is a member of the NIHR Editorial Board and the NIHR Journals Library Editorial Group. He was previously Director of the Wessex Institute and Head of the NIHR Evaluation, Trials and Studies Coordinating Centre.

Published October 2015

DOI: 10.3310/hta19780

This report should be referenced as follows:

Chakravarthy U, Harding SP, Rogers CA, Downes S, Lotery AJ, Dakin HA,et al. A randomised controlled trial to assess the clinical effectiveness and cost-effectiveness of alternative treatments to Inhibit VEGF in Age-related choroidal Neovascularisation (IVAN).Health Technol Assess 2015;19(78).

(6)

(7)

ISSN 1366-5278 (Print) ISSN 2046-4924 (Online) Impact factor: 5.116

Health Technology Assessmentis indexed in MEDLINE, CINAHL, EMBASE, The Cochrane Library and the ISI Science Citation Index. This journal is a member of and subscribes to the principles of the Committee on Publication Ethics (COPE) (www.publicationethics.org/). Editorial contact: [email protected]

The full HTA archive is freely available to view online at www.journalslibrary.nihr.ac.uk/hta. Print-on-demand copies can be purchased from the report pages of the NIHR Journals Library website: www.journalslibrary.nihr.ac.uk

Criteria for inclusion in theHealth Technology Assessmentjournal

Reports are published inHealth Technology Assessment(HTA) if (1) they have resulted from work for the HTA programme, and (2) they are of a sufficiently high scientific quality as assessed by the reviewers and editors.

Reviews inHealth Technology Assessmentare termed‘systematic’when the account of the search appraisal and synthesis methods (to minimise biases and random errors) would, in theory, permit the replication of the review by others.

HTA programme

The HTA programme, part of the National Institute for Health Research (NIHR), was set up in 1993. It produces high-quality research information on the effectiveness, costs and broader impact of health technologies for those who use, manage and provide care in the NHS.

‘Health technologies’are broadly defined as all interventions used to promote health, prevent and treat disease, and improve rehabilitation and long-term care.

The journal is indexed in NHS Evidence via its abstracts included in MEDLINE and its Technology Assessment Reports inform National Institute for Health and Care Excellence (NICE) guidance. HTA research is also an important source of evidence for National Screening Committee (NSC) policy decisions.

For more information about the HTA programme please visit the website: http://www.nets.nihr.ac.uk/programmes/hta This report

The research reported in this issue of the journal was funded by the HTA programme as project number 07/36/01. The contractual start date was in July 2007. The draft report began editorial review in January 2014 and was accepted for publication in November 2014. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’report and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health. If there are verbatim quotations included in this publication the views and opinions expressed by the

interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health.

© Queen’s Printer and Controller of HMSO 2015. This work was produced by Chakravarthyet al.under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

(8)

Professor Tom Walley Director, NIHR Evaluation, Trials and Studies and Director of the HTA Programme, UK

NIHR Journals Library Editors

Professor Ken Stein Chair of HTA Editorial Board and Professor of Public Health, University of Exeter Medical School, UK

Professor Andree Le May Chair of NIHR Journals Library Editorial Group (EME, HS&DR, PGfAR, PHR journals)

Dr Martin Ashton-Key Consultant in Public Health Medicine/Consultant Advisor, NETSCC, UK

Professor Matthias Beck Chair in Public Sector Management and Subject Leader (Management Group), Queen’s University Management School, Queen’s University Belfast, UK

Professor Aileen Clarke Professor of Public Health and Health Services Research, Warwick Medical School, University of Warwick, UK

Dr Tessa Crilly Director, Crystal Blue Consulting Ltd, UK

Dr Peter Davidson Director of NETSCC, HTA, UK

Ms Tara Lamont Scientific Advisor, NETSCC, UK

Professor Elaine McColl Director, Newcastle Clinical Trials Unit, Institute of Health and Society, Newcastle University, UK

Professor William McGuire Professor of Child Health, Hull York Medical School, University of York, UK

Professor Geoffrey Meads Professor of Health Sciences Research, Faculty of Education, University of Winchester, UK

Professor John Norrie Health Services Research Unit, University of Aberdeen, UK

Professor John Powell Consultant Clinical Adviser, National Institute for Health and Care Excellence (NICE), UK

Professor James Raftery Professor of Health Technology Assessment, Wessex Institute, Faculty of Medicine, University of Southampton, UK

Dr Rob Riemsma Reviews Manager, Kleijnen Systematic Reviews Ltd, UK

Professor Helen Roberts Professor of Child Health Research, UCL Institute of Child Health, UK

Professor Helen Snooks Professor of Health Services Research, Institute of Life Science, College of Medicine, Swansea University, UK

Professor Jim Thornton Professor of Obstetrics and Gynaecology, Faculty of Medicine and Health Sciences, University of Nottingham, UK

Please visit the website for a list of members of the NIHR Journals Library Board: www.journalslibrary.nihr.ac.uk/about/editors

(9)

Abstract

A randomised controlled trial to assess the clinical

effectiveness and cost-effectiveness of alternative

treatments to Inhibit VEGF in Age-related choroidal

Neovascularisation (IVAN)

Usha Chakravarthy,

1*

_{Simon P Harding,}

2

_{Chris A Rogers,}

3

Susan Downes,

4

_{Andrew J Lotery,}

5

_{Helen A Dakin,}

6

_{Lucy Culliford,}

3

Lauren J Scott,

3

_{Rachel L Nash,}

3

_{Jodi Taylor,}

3

_{Alyson Muldrew,}

1

Jayashree Sahni,

2

_{Sarah Wordsworth,}

6

_{James Raftery,}

7

_{Tunde Peto}

8

and Barnaby C Reeves

3

_{for the IVAN Investigators}

†

1_{Centre for Experimental Medicine, Institute of Clinical Science, Queen}’_{s University Belfast,}

Belfast, UK

2_{Department of Eye and Vision Science, Institute of Ageing and Chronic Disease,}

University of Liverpool, Liverpool, UK

3_{Clinical Trials and Evaluation Unit, School of Clinical Sciences, University of Bristol, Bristol, UK}

4_{Oxford University Hospitals NHS Trust, Oxford, UK}

5_{Clinical and Experimental Sciences, Faculty of Medicine, University of Southampton,}

Southampton, UK

6_{Health Economic Research Centre, Nuffield Department of Population Health,}

University of Oxford, Oxford, UK

7_{Wessex Institute, University of Southampton, Southampton, UK}

8_{National Institute for Health Research (NIHR) Biomedical Research Centre at Moorfields Eye}

Hospital NHS Foundation Trust and UCL Institute of Ophthalmology, London, UK

*Corresponding author [email protected] †IVAN Investigators are listed inAppendix 1.

Background:Bevacizumab (Avastin®_{, Roche), which is used in cancer therapy, is the}‘_parent’_molecule from which ranibizumab (Lucentis®_{, Novartis) was derived for the treatment of neovascular age-related} macular degeneration (nAMD). There were reports in the literature on the effectiveness of bevacizumab in treating nAMD, but no trials. The cost per dose of bevacizumab is about 5–10% that of ranibizumab. This trial was a head-to-head comparison of these two drugs.

Objective:To compare the clinical effectiveness and cost-effectiveness of ranibizumab and bevacizumab, and two treatment regimens, for nAMD.

Design:Multicentre, factorial randomised controlled trial with within-trial cost–utility and cost-minimisation analyses from the perspective of the UK NHS. Participants, health professionals and researchers were masked to allocation of drug but not regimen. Computer-generated random allocations to combinations of ranibizumab or bevacizumab, and continuous or discontinuous regimen, were stratified by centre, blocked and concealed.

(10)

Participants:Patients≥50 years old with active nAMD in the study eye with best corrected distance visual acuity (BCVA)≥25 letters measured on a Early Treatment of Diabetic Retinopathy Study (ETDRS) chart. Previous treatment for nAMD, long-standing disease, lesion diameter>6000 µm, thick blood at the fovea and any other confounding ocular disease were exclusion criteria. One eye per participant was studied; the fellow eye was treated according to usual care, if required.

Interventions:Ranibizumab and bevacizumab were procured commercially. Doses were ranibizumab 0.5 mg or bevacizumab 1.25 mg. The repackaged bevacizumab was quality assured. All participants were treated at visits 0, 1 and 2. Participants randomised to the continuous regimen were treated monthly thereafter. Participants randomised to the discontinuous regimen were not retreated after visit 2 unless pre-specified criteria for active disease were met. If retreatment was needed, monthly injections over 3 months were mandated.

Main outcome measures:The primary outcome was BCVA. The non-inferiority margin was 3.5 letters. Secondary outcomes were contrast sensitivity; near visual acuity; reading index; neovascular lesion morphology; generic and disease-specific patient-reported outcomes, including macular disease-specific quality of life; survival free from treatment failure; resource use; quality-adjusted life-years (QALYs); and development of new geographic atrophy (GA) (outcome added during the trial). Results are reported for the study eye, except for patient-reported outcomes.

Results:Between 27 March 2008 and 15 October 2010, 610 participants were allocated and treated (314 ranibizumab, 296 bevacizumab; at 3 months, 305 continuous, 300 discontinuous). After 2 years, bevacizumab was neither non-inferior nor inferior to ranibizumab [–1.37 letters, 95% confidence interval (CI)–3.75 to+1.01 letters] and discontinuous treatment was neither non-inferior nor inferior to continuous treatment (–1.63 letters, 95% CI–4.01 to+0.75 letters). Lesion thickness at the fovea was similar by drug [geometric mean ratio (GMR) 0.96, 95% CI 0.90 to 1.03;p=0.24] but 9% less with continuous treatment (GMR 0.91, 95% CI 0.85 to 0.97;p=0.004). Odds of developing new GA during the trial were similar by drug [odds ratio (OR) 0.87, 95% CI 0.61 to 1.25;p=0.46] but significantly higher with continuous

treatment (OR 1.47, 95% CI 1.03 to 2.11;p=0.033). Safety outcomes did not differ by drug but mortality was lower with continuous treatment (OR 0.47, 95% CI 0.22 to 1.03;p=0.05). Continuous ranibizumab cost £3.5M per QALY compared with continuous bevacizumab; continuous bevacizumab cost £30,220 per QALY compared with discontinuous bevacizumab. These results were robust in sensitivity analyses.

Conclusions:Ranibizumab and bevacizumab have similar efficacy. Discontinuing treatment and restarting when required results in slightly worse efficacy. Safety was worse with discontinuous treatment, although new GA developed more often with continuous treatment. Ranibizumab is not cost-effective, although it remains uncertain whether or not continuous bevacizumab is cost-effective compared with discontinuous bevacizumab at £20,000 per QALY threshold. Future studies should focus on the ocular safety of the two drugs, further optimisation of treatment regimens and criteria for stopping treatment.

Trial registration:Current Controlled Trials ISRCTN92166560.

Funding:This project was funded by the NIHR Health Technology Assessment programme and will be published in full inHealth Technology Assessment; Vol. 19, No. 78. See the NIHR Journals Library website for further project information.

(11)

List of tables

TABLE 1 Factorial study design 5

TABLE 2 Definitions of measures of lesion morphology and metrics from colour

fundus, fluorescein angiographic and OCT images 10

TABLE 3 Total sample size estimates for objective I 14

TABLE 4 Schedule of data collection 16

TABLE 5 Summary of methods used in economic evaluation 24

TABLE 6 Methods used to calculate consultation costs 28

TABLE 7 Assumptions made about the frequency of injection and monitoring

consultations within the costing analysis 29

TABLE 8 Withdrawals 40

TABLE 9 Protocol deviations 42

TABLE 10 Patient demography and past history 46

TABLE 11 Best corrected distance VA at baseline, 1 year and 2 years 50

TABLE 12 Secondary measures of visual function at baseline, 1 year and 2 years 54

TABLE 13 Morphological characteristics present or absent at baseline, 1 year and

2 years 66

TABLE 14 Morphological characteristics quantified at baseline, 1 year and 2 years 70

TABLE 15 Optical coherence tomography metrics at baseline, 1 year and 2 years 72

TABLE 16 Vascular events, HF and death 78

TABLE 17 Causes of death by drug and regimen combination 80

TABLE 18 Serious adverse events by system organ class 81

TABLE 19 Ocular SAEs in the non-study eye 92

TABLE 20 Serious adverse events classified as possibly, probably or definitely

related to treatment 93

TABLE 21 Adverse events by system organ class 95

(16)

TABLE 23 The EQ-5D utility, macular disease-specific quality of life and

treatment satisfaction scores 112

TABLE 24 Blood pressure 115

TABLE 25 Number and percentage of patients in each group with a decrease of

≥15 letters between consecutive visits during the trial 118

TABLE 26 Number and percentage of patients in each group reporting

worsening angina during the trial 118

TABLE 27 Cost of consultations for intravitreal injections and evaluating

response to treatment 119

TABLE 28 Comparison of quantities of each type of NHS resource uses used by

the average patient in each study arm over the 2-year study period 122

TABLE 29 Breakdown of costs and QALYs per patient by quarters 123

TABLE 30 Results of base-case economic evaluation: mean QALYs, costs and net

benefits per patient over the 2-year trial period 124

TABLE 31 Results of sensitivity 127

TABLE 32 Further details on protocol deviations 173

TABLE 33 Visit profile by centre 174

TABLE 34 Bands of VA at baseline and 2 years 176

TABLE 35 Distribution of responses to individual EQ-5D questions at baseline

and 2 years 180

TABLE 36 Variables included in mixed models 187

TABLE 37 Results of the best-fitting models for EQ-5D and HUI3 189

TABLE 38 Unit costs used in the economic evaluation 198

TABLE 39 Cost per admission 205

(17)

List of figures

FIGURE 1 Criteria for treatment failure/restarting treatment 7

FIGURE 2 Treatment over time by treatment regimen 8

FIGURE 3 New GA assessment diagram 13

FIGURE 4 How QALYs were calculated 25

FIGURE 5 Contracts and monitoring 36

FIGURE 6 Finance flow chart 36

FIGURE 7 Data flow to and from the Trial Coordinating Centre (TCC) 36

FIGURE 8 Recruitment rate 37

FIGURE 9 Number of participants recruited by centre 38

FIGURE 10 Flow of participants 39

FIGURE 11 Time from randomisation to withdrawal (excluding deaths) 40

FIGURE 12 Distribution of the number of missed visits per visit (a) by drug and

(b) by treatment regimen 43

FIGURE 13 Distribution of the number of visits attended per participant

(a) by drug and (b) by treatment regimen 44

FIGURE 14 Distribution of the number of injections received per participant

FIGURE 15 Best corrected distance visual acuity in the study eye by visit

FIGURE 16 Change from baseline in BCVA in the study eye by visit (a) by drug

and (b) by treatment regimen 52

FIGURE 17 Difference in BCVA at 2 years 53

FIGURE 18 Near visual acuity in the study eye by visit (a) by drug and

FIGURE 19 Reading index in the study eye by visit (a) by drug and

FIGURE 20 Contrast sensitivity in the study eye by visit (a) by drug and

(18)

FIGURE 21 Difference in secondary measures of visual function at 2 years

FIGURE 22 Time to first treatment failure (a) by drug and (b) by treatment regimen 60

FIGURE 23 Best corrected distance visual acuity in the study eye at 2 years by

subgroup (a) by drug and (b) by treatment regimen 61

FIGURE 24 Meta-analysis of change in BVCA from baseline (a) by drug and

FIGURE 25 Secondary morphological outcomes at 2 years (a) by drug and

FIGURE 26 Meta-analysis of change in total retinal thickness at the fovea from

baseline (a) by drug and (b) by treatment regimen 75

FIGURE 27 Meta-analysis of GA (a) by drug and (b) by treatment regimen 76

FIGURE 28 Safety outcomes at 2 years (a) by drug and (b) by treatment regimen 79

FIGURE 29 Safety outcomes classified by MedDRA system organ class at 2 years

FIGURE 30 Meta-analysis of safety outcomes (a) by drug and (b) by treatment

regimen 107

FIGURE 31 Meta-analysis of gastrointestinal SAEs (a) by drug and (b) by treatment

regimen 109

FIGURE 32 The EQ-5D utility, macular disease-specific quality of life and

treatment satisfaction scores (a) by drug and (b) by treatment regimen 113

FIGURE 33 The EQ-5D utility, macular disease-specific quality of life and treatment satisfaction scores for subgroups of participants with vision in the

study eye>5 letters and≤5 letters better (a) by drug and (b) by treatment frequency 114 FIGURE 34 Systolic blood pressure by visit (a) by drug and (b) by treatment

regimen 116

FIGURE 35 Diastolic blood pressure by visit (a) by drug and (b) by treatment

regimen 117

FIGURE 36 Distribution of consultation costs across IVAN centres 119

FIGURE 37 Breakdown of costs for each treatment 126

FIGURE 38 Differences in costs for ranibizumab against the difference in QALYs

for (a) continuous treatment and (b) discontinuous treatment 131

FIGURE 39 Pairwise cost-effectiveness acceptability curves for ranibizumab

vs. bevacizumab 132

(19)

FIGURE 40 Incremental costs and outcomes for continuous vs. discontinuous

treatment for (a) ranibizumab and (b) bevacizumab 133

FIGURE 41 Pairwise cost-effectiveness acceptability curves for comparison

between continuous and discontinuous treatment 134

FIGURE 42 Total costs against total QALYs for each of the four treatments 135

FIGURE 43 Cost-effectiveness acceptability curves for multiple comparisons based

on comparisons between all four treatment strategies 135

FIGURE 44 Two-way sensitivity analysis showing the impact of drug cost on the incremental cost-effectiveness of continuous ranibizumab vs. continuous

bevacizumab 137

FIGURE 45 Binocular vision by visit (a) by drug and (b) by treatment regimen 177

FIGURE 46 Best corrected distance visual acuity in the study eye at 2 years by subgroup, with baseline BCVA in study eye better than fellow eye vs. the same

or worse (a) by drug and (b) by treatment regimen 178

FIGURE 47 Best corrected distance visual acuity in the study eye at 2 years by subgroup with haemorrhage in the study eye at baseline present vs. absent

FIGURE 48 Possible assumptions for estimating the EQ-5D profile around SAEs 185

FIGURE 49 The effect of model coefficients on EQ-5D utility for a hypothetical patient who experienced a cardiovascular SAE on day 90, an ocular SAE on day 150, an unexpected SAE on day 205, a second cardiovascular SAE on day 330

and a cancer SAE on day 629 190

FIGURE 50 Illustration of the methods used to infer the EQ-5D profile and calculate QALYs for a hypothetical patient with six SAEs and three unscheduled

measurements 193

FIGURE 51 Assumptions made in the base-case analysis and the three main

(20)

(21)

List of abbreviations

AE adverse event

AIC Akaike information criterion

AMD age-related macular degeneration

ANCHOR ANti-vascular endothelial growth factor antibody for the treatment of predominantly Classic

CHORoidal neovascularization in age-related macular degeneration

ATE arterial thrombotic event

BCVA best corrected distance visual acuity

BRAMD comparison of bevacizumab

(Avastin) and ranibizumab (Lucentis) in exudative age-related macular degeneration

CARF Central Angiographic Resource

Facility

CATT Comparison of Age-related

macular degeneration Treatment Trials

CCS Canadian Cardiovascular Society

CI confidence interval

CMA cost-minimisation analysis

CNV choroidal neovascularisation

CRF case report form

CTEU Clinical Trials and Evaluation Unit

CUA cost–utility analysis

CVD cardiovascular disease

df degree of freedom

DMSC Data Monitoring and Safety

Committee

DNA deoxyribonucleic acid

DVT deep-vein thrombosis

EQ-5D European Quality of Life-5

Dimensions

ETDRS Early Treatment of Diabetic

Retinopathy Study

FFA fundus fluorescein angiography

GA geographic atrophy

GEFAL French Evaluation Group Avastin

versus Lucentis

GMR geometric mean ratio

GP general practitioner

HF heart failure

HR hazard ratio

HRG Healthcare Resource Group

HRQoL health-related quality of life HUI3 Health Utilities Index, version 3 ICER incremental cost-effectiveness ratio

IMP investigational medicinal product

IOP intraocular pressure

IQR interquartile range

IVAN Inhibit VEGF in Age-related

choroidal Neovascularisation

KMSA Kaplan–Meier sample averaging

logMAR log(minimum angle of resolution)

LUCAS LUcentis Compared to Avastin

Study

M cost per monitoring consultation

excluding FFA and any resources associated with the intravitreal injection

MacDQoL Macular disease Dependent Quality

of Life

MacTSQ Macular disease Treatment

Satisfaction Questionnaire

MANTA Multicenter ANti-VEGF Trial in

Austria

MARINA Minimally classic/occult trial of the Anti-VEGF antibody Ranibizumab In the treatment of Neovascular Age-related macular degeneration

(22)

MedDRA Medical Dictionary for Regulatory Activities

MHRA Medicines and Healthcare products

Regulatory Agency

MI myocardial infarction

nAMD neovascular age-related macular

degeneration

NICE National Institute for Health and

Care Excellence

NIHR National Institute for Health

Research

NVA near visual acuity

NYHA New York Heart Association

OCT optical coherence tomography

OLS ordinary least squares

OR odds ratio

PDT photodynamic therapy

PED pigment epithelial detachment

PIER a study of rhuFAB V2 [ranibizumab]

in subjects with subfoveal choroidal neovascularization secondary to age-related macular degeneration

prn pro re nata (as needed)

QALY quality-adjusted life-year

QUB Queen’s University Belfast

RAP retinal angiomatous proliferation

RCT randomised controlled trial

RPE retinal pigment epithelium

SAE serious adverse event

SAP statistical analysis plan

SD standard deviation

SE standard error

SRF subretinal fluid

TCC Trial Coordinating Centre

TSC Trial Steering Committee

VA visual acuity

VAT value added tax

VEGF vascular endothelial growth factor

VIEW VEGF Trap-Eye: Investigation of

(1 and 2) Efficacy and Safety in Wet AMD

VISION pegaptanib for neovascular

age-related macular degeneration

VPDT Verteporfin PhotoDynamic Therapy

cohort study

(23)

Plain English summary

T

he study aimed to compare two drugs (bevacizumab and ranibizumab), and two dosing intensities (injections monthly or intermittently), for treating wet age-related macular degeneration (AMD). Twenty-three hospitals took part. Broadly, we sought to include patients with wet AMD in the study eye, who were older than 50 years. Each participant was allocated one of the drugs and one of the dosing intensities, creating four combinations. The main outcome was the number of letters read on an eye chart using the study eye. Other aspects of eyesight, the appearance of the affected part of the eye, possible harms of treatment, patient-reported outcomes, and resources used to treat participants were measured.

The average improvement in eyesight was very similar with either drug. Monthly treatment was slightly better than intermittent treatment. Quality of life and treatment satisfaction did not differ by drug or dosing intensity. When our results were combined with those of other trials, there was no difference in eyesight outcomes between drugs, but giving treatment monthly was slightly better. Hospitalisations and deaths during the trial occurred equally often with either drug but less often with monthly than intermittent treatment. Ranibizumab was 15 times more expensive than bevacizumab.

(24)

(25)

Scientific summary

Introduction

Neovascular age-related macular degeneration (nAMD) is a bilateral condition causing severe central vision impairment. Ranibizumab (Lucentis®_{, Novartis), an antibody to vascular endothelial growth factor (VEGF),} is an effective treatment. Bevacizumab (Avastin®_{, Roche), the parent molecule for ranibizumab, is licensed} for other indications but not nAMD. It was identified as having similar benefits but at much lower cost. When the alternative treatments to the Inhibit VEGF in Age-related choroidal Neovascularisation (IVAN) trial was conceived, there was no systematic review of VEGF inhibitors to treat nAMD, no head-to-head comparison of the two drugs and no data on minimum treatment frequency.

Objectives

The trial had three objectives. To estimate:

i. the effectiveness of bevacizumab compared with ranibizumab

ii. the effectiveness of discontinuous versus continuous treatment regimens, with criteria for restarting treatment when required in patients receiving discontinuous treatment

iii. the cost-effectiveness of the alternative treatment strategies outlined above.

Methods

Study design

Multicentre, randomised, controlled factorial trial and within-trial economic evaluation, comparing the two drugs and two treatment regimens.

Settings and participants

Patients were recruited from UK NHS hospitals. Adults of≥50 years of age, newly referred with nAMD in either eye, a best corrected distance visual acuity (BCVA) of≥25 letters and a foveal neovascular lesion were eligible. Previous treatment for nAMD, fibrosis>50% of the total lesion, greatest linear diameter of

>6000 µm, thick blood involving the centre of the fovea, other active ocular disease causing vision loss or≥8 dioptres of myopia in the proposed study eye were exclusion criteria.

Interventions

Participants were allocated to one of four combinations: intravitreal injection of ranibizumab (0.5 mg) or bevacizumab (1.25 mg), and‘continuous’or‘discontinuous’treatment regimens. All participants attended monthly and were treated at visits 0, 1 and 2. Participants randomised to the continuous regimen were treated monthly thereafter; participants randomised to the discontinuous regimen were treated only if prespecified clinical and optical coherence tomography (OCT) criteria for active disease were met. If retreatment was initiated, three further doses at monthly intervals were mandated.

Randomisation

Randomisation was stratified by centre and blocked. Allocations were computer generated and concealed.

Masking

(26)

Outcomes

The primary outcome was BCVA, assessed at baseline 3, 6, 12, 18 and 24 months. The primary end point was after 2 years of follow-up.

Secondary outcomes were:

(a) Contrast sensitivity, near visual acuity (NVA) and reading index.

(b) Lesion morphology [from colour fundus photography, fundus fluorescein angiography (FFA) and OCT]. (c) Two generic health status measures, European Quality of Life-5 Dimensions (EQ-5D) and Health Utilities

Index version 3 (HUI3), and two macular-disease specific instruments: the MacDQoL (Macular disease Dependent Quality of Life: measuring the impact of macular disease on quality of life) and MacTSQ (Macular disease Treatment Satisfaction Questionnaire: measuring satisfaction with treatment for macular disease).

(d) Survival free from treatment failure.

(e) Resource use and quality-adjusted life-years (QALYs).

(f) Adverse events (AEs); the primary safety outcome was an arterial thrombotic event (ATE) or hospital admission for heart failure.

(g) Development of geographic atrophy (GA) during the trial.

Follow-up

Participants were followed monthly for 2 years. Colour photographs and OCTs were captured every 3 months, and FFA at baseline, 12 and 24 months. Participants completed the EQ-5D and HUI3 at visits 0, 3, 12 and 24, and when a serious adverse event (SAE) had occurred since the previous visit. MacDQoL and MacTSQ questionnaires were administered by telephone after visits 3, 12 and 24.

Sample size

We aimed to test non-inferiority hypotheses about visual function, with non-inferiority margins of 3 or 4 letters and analyses using one or two BCVA measurements, adjusted for baseline BCVA. Other assumptions were:

l no interaction between drug and treatment regimen l BCVA standard deviation (SD)=14 letters

l 90% power, 2.5% significance

l correlation between baseline and follow-up BCVA=0.5, and between follow-up BCVA=0.8.

For a 3-letter non-inferiority margin, 600 participants were required.

A prespecified interim analysis was undertaken after participants had been followed for 1 year.

Statistical analyses

Intention-to-treat analyses were carried out. Continuously scaled outcomes at multiple time points were analysed using linear mixed-effects methods. Binary outcomes were analysed using logistic regression, only if≥10 participants experienced the outcome. Time to first treatment failure was analysed using Cox proportional hazards regression.

Effect estimates were reported separately by treatment regimen if the interaction of drug and treatment regimen reached statistical significance; otherwise, main effects of drug and treatment regimen were reported. No adjustment was made for multiple testing. Pre-planned subgroup analyses were tested by adding subgroup-by-treatment interactions. Fixed-effects meta-analyses combining the results of IVAN with other head-to-head trials were undertaken to place the trial findings within the context of existing evidence.

(27)

Economic evaluation

A within-trial economic evaluation was conducted to assess the incremental cost and cost-effectiveness of discontinuous and continuous treatment using bevacizumab and ranibizumab from the cost perspective of the NHS and the health perspective of participants. Health-care resource use was collected for all trial participants. The analysis included the cost of study medication; drug administration/monitoring consultations; any concomitant medication, ambulatory consultations and hospitalisations for‘related’AEs or SAEs. Costs and QALYs were discounted at 3.5% per annum.

Results

Patient screening

In total, 693 patients were screened: 65 excluded and 628 randomised.

Recruitment

Recruitment occurred between 27 March 2008 and 15 October 2010, with the last follow-up on 7 November 2012. Five of 628 randomised participants were subsequently found to be ineligible and 13 were not treated, leaving 610 who received at least one injection in the IVAN study cohort.

Withdrawals

Sixty participants withdrew. The most common reasons for withdrawal were illness preventing attendance and occurrence of a SAE.

Follow-up

In total, 525 of 610 participants completed the trial, with 87% of all scheduled visits attended. Missed visits were distributed similarly across groups.

Trial cohort

Mean age was 77.7 years (SD 7.4 years); 244 (40%) participants were male, 64% were current or past smokers and 19% had a history of dyspnoea. Characteristics were similar between groups, although more participants allocated to bevacizumab than ranibizumab had angina (17% vs. 11%).

Treatment received

The number of injections administered was similar by drug (ranibizumab: median 18; bevacizumab: median 19). More injections were given with continuous than discontinuous treatment (medians 23 vs. 13).

Success of masking

Ophthalmologists and participants reported not knowing which drug participants were receiving on>97% of 3-, 12- and 24-month visits.

Unmasking

Unmasking was not required.

Primary outcome: best corrected distance visual acuity

(28)

A meta-analysis of changes in BCVA from baseline in seven trials showed that bevacizumab was

statistically non-inferior to ranibizumab (–0.38 letters, 95% CI–1.47 to+0.70 letters;p=0.49). Only CATT (Comparison of Age-related macular degeneration Treatment Trials) and IVAN compared treatment regimens; their combined data showed that discontinuous treatment was significantly inferior to continuous treatment (–2.23 letters, 95% CI–3.93 to–0.53 letters;p=0.010).

Secondary measures of visual function at 2 years

Near visual acuity did not differ by drug [geometric mean ratio (GMR)=0.94, 95% CI 0.85 to 1.04; p=0.23] but was better with continuous treatment (GMR=0.90, 95% CI 0.82 to 0.99;p=0.04). Reading index was similar by drug and regimen. Contrast sensitivity did not differ significantly by drug, but was better with continuous treatment (mean difference=–1.07, 95% CI–1.90 to–0.25;p=0.011). Lesion morphology at 2 years

There were no significant differences by drug or treatment regimen for dye leakage on FFA but fewer participants treated with ranibizumab had fluid on OCT (50% vs. 59%;p=0.065) and when treated continuously [45% vs. 63%, odds ratio (OR)=0.47, 95% CI 0.33 to 0.67;p<0.001].

Total thickness, and retinal plus subfoveal fluid thickness, at the fovea did not differ by drug. However, they were 9% and 8% less thick for continuous treatment (GMR=0.91, 95% CI 0.85 to 0.97;p=0.004; GMR=0.92, 95% CI 0.84 to 1.00;p=0.046, respectively).

A meta-analysis of changes in total retinal thickness at the fovea suggested a non-statistically significant difference in favour of ranibizumab (p=0.12). The combined CATT and IVAN data showed a statistically significant difference in favour of continuous treatment (p=0.001).

New GA during the trial developed in 30% of participants. There was no difference by drug (28% for ranibizumab vs. 31% for bevacizumab: OR=0.87, 95% CI 0.61 to 1.25;p=0.46) but GA developed significantly more often in the continuous group (34% vs. 26%, OR=1.47, 95% CI 1.03 to 2.11; p=0.03). This finding was confirmed in a meta-analysis (OR=1.56, 95% CI 1.20 to 2.03;p=0.001). Adverse events

Overall, 171 participants had one or more SAEs, of whom 30 died. The primary safety end point frequency did not differ significantly by drug (OR=1.69, 95% CI 0.80 to 3.57;p=0.16) or regimen (OR=0.56, 95% CI 0.27 to 1.19;p=0.13). Deaths were split equally by drug but occurred more frequently in the discontinuous than continuous group (20 vs. 10, OR=0.47, 95% CI 0.22 to 1.03;p=0.05).

Gastrointestinal SAEs appeared to occur more frequently with bevacizumab. The percentages of patients having any systemic SAE were similar by drug (ranibizumab 26%, bevacizumab 27%) and treatment regimen (continuous 24%, discontinuous 29%); 39 non-ocular SAEs were classified as possibly, probably or definitely related to treatment.

Meta-analyses of safety outcomes showed no differences by drug for deaths or ATEs but a significantly increased risk of any systemic SAE for bevacizumab (OR=0.77, 95% CI 0.64 to 0.92;p=0.004). The comparison by treatment regimen showed increased risks of death (OR=0.49, 95% CI 0.27 to 0.86;p=0.014) and any systemic SAE (OR=0.81, 95% CI 0.65 to 1.01;p=0.063) with discontinuous treatment.

A post hoc meta-analysis of gastrointestinal SAEs showed a significantly increased risk of gastrointestinal SAEs in the bevacizumab group (OR=0.53, 95% CI 0.33 to 0.85;p=0.009) but no difference by treatment regimen (OR=0.88, 95% CI 0.44 to 1.78;p=0.73).

(29)

Patient-reported outcomes

The MacDQoL and MacTSQ scores and EQ-5D utilities were very similar at 1 and 2 years, both by drug and treatment regimen (p≥0.23).

Economic evaluation

All four groups accrued an average of 1.6 QALYs over 2 years, with no differences by drug or treatment regimen (p≥0.381). Total 2-year costs ranged from £3002 per patient (95% CI £2601 to £3403) for

discontinuous bevacizumab to £18,590 per patient (95% CI £18,258 to £18,922) for continuous ranibizumab.

Ranibizumab was significantly more costly than bevacizumab (p<0.001), costing an additional £14,989 per patient (95% CI £14,522 to £15,456) for continuous treatment and £8498 per patient (95% CI £7700 to £9295) for discontinuous treatment. As QALY differences were negligible, continuous ranibizumab cost £3.5M per QALY gained compared with continuous bevacizumab. Bootstrapping demonstrated that we can be>99.99% confident that continuous ranibizumab is poor value for money compared with discontinuous ranibizumab at a £20,000 per QALY ceiling ratio.

Patients receiving continuous versus discontinuous bevacizumab accrued higher total costs (£599, 95% CI £91 to £1107;p=0.021) but also accrued non-significantly more QALYs (mean difference: 0.020, 95% CI

–0.032 to 0.071;p=0.452). Continuous bevacizumab therefore cost £30,220 per QALY gained compared with discontinuous bevacizumab. However, this finding was substantially uncertain, with a 37% chance that continuous bevacizumab is cost-effective at a £20,000 per QALY ceiling ratio.

Discussion

Main findings: study results

Comparisons by drug and treatment regimen for BCVA were inconclusive. However, the BCVA meta-analysis showed that bevacizumab is non-inferior to ranibizumab and that discontinuous treatment is significantly inferior to continuous treatment. Secondary visual function and lesion morphology outcomes were consistent with the BCVA meta-analyses. New GA developed more often with continuous than discontinuous treatment.

Deaths and SAEs did not differ by drug. The meta-analyses of safety data by drug showed a significant increase in the risk of any systemic SAE, and SAEs classified as gastrointestinal, with bevacizumab. There were twice as many deaths in IVAN with discontinuous treatment, a finding confirmed in a meta-analysis. The odds of any systemic SAE by treatment regimen also tended to favour continuous treatment.

The economic evaluation demonstrated that ranibizumab is not cost-effective compared with bevacizumab. If hospitals in England were to switch from discontinuous ranibizumab to discontinuous bevacizumab, the NHS could save at least £102M per year [including 20% value added tax (VAT)]. Discontinuous bevacizumab is likely to be the most cost-effective treatment strategy evaluated in IVAN. However, this finding is substantially uncertain, with a 37% chance that continuous bevacizumab is cost-effective.

Strengths and limitations

The IVAN trial should directly inform the use of anti-VEGF drugs in the NHS. Secondary visual function outcomes supported the BCVA findings. Masking of the allocated drug and adherence to allocations of drug and treatment regimen were excellent. Retention of the elderly participants was good, with only 10% withdrawing. A detailed health-economic evaluation was carried out.

(30)

Lessons for the future

Because the trial was considered to be‘high risk,’the IVAN trial design was modelled on commercial trials. To better understand outcomes in a clinical setting IVAN could have been more pragmatic; for example, we could have recruited both eyes, if both were eligible, and managed them according to same allocation.

In future, economic models to estimate the cost-effectiveness of interventions for nAMD should use robust associations between VA and utility estimated from large data sets using methods recommended by the National Institute for Health and Care Excellence. Our findings about the development of new GA highlight that these models may need to consider a long time horizon.

Combination therapies are being investigated with the aim of reducing treatment frequency. Our finding of possible risks of discontinuous treatment highlights the importance of careful monitoring of SAEs with such treatment regimens.

Conclusion

The IVAN trial and meta-analyses of data from other trials show that the choice of anti-VEGF treatment strategy is less straightforward than previously thought. Bevacizumab and ranibizumab have similar efficacy. Continuous treatment avoids the need to monitor disease activity on every visit, with slightly better functional outcomes. Our economic evaluation showed that ranibizumab represents poor value for money, discontinuous bevacizumab is probably better value for money than continuous bevacizumab, and monthly treatment with ranibizumab is unaffordable for publicly funded health systems.

Implications for health care Findings from the IVAN trial:

l support using bevacizumab, which was non-inferior to ranibizumab, for both efficacy and safety l highlight that economic models should use robust associations between VA and health-related quality

of life, estimated from large data sets and adopting a long time horizon

l identify the need to monitor the frequency of SAEs with combination treatment regimens designed to reduce treatment frequency.

Recommendations for research Research is needed to investigate:

l different models of service provision l when treating nAMD may be futile

l the long-term consequences of anti-VEGF treatment l reasons for poorer safety with discontinuous treatment.

Trial registration

This trial is registered as ISRCTN92166560.

Funding

Funding for this study was provided by the Health Technology Assessment programme of the National Institute for Health Research.

(31)

Chapter 1 Introduction

Background and rationale

Introduction

The onset of neovascularisation in age-related macular degeneration (AMD) is accompanied by central distortion and blurring which, when left untreated, intensifies into a dense central scotoma.1_{These visually} disabling effects of neovascular age-related macular degeneration (nAMD) are monitored by measuring distance visual acuity (VA), which is a surrogate for central visual function; a drop of≥15 letters in the number of letters read on a logMAR [log(minimum angle of resolution)] letter chart (equivalent to the loss of three lines of letters) is considered to be a visually significant event.2_{Until 2005, all treatments that were} considered beneficial in the management of nAMD merely limited VA loss relative to untreated control subjects or natural history.3,4_{Thus, at that time successful treatment was defined in terms of slowing} down the rate of VA loss. The most effective treatment for nAMD was verteporfin (Visudyne, Novartis) photodynamic therapy (PDT) which resulted in a reduction of the proportion of patients who suffered a 15-letter loss [three lines on the Early Treatment of Diabetic Retinopathy Study (ETDRS) VA chart] from 69% to 54%.3,5_{The most impressive change that occurred around this time was the introduction of} ranibizumab (Lucentis®_{, Novartis). Ranibizumab is a monoclonal antibody to vascular endothelial growth} factor (VEGF), which is a potent mitogen and inducer of permeability in blood vessels.6_{Two pivotal} ANCHOR (ANti-vascular endothelial growth factor antibody for the treatment of predominantly Classic CHORoidal neovascularization in age-related macular degeneration) and MARINA (Minimally classic/occult trial of the Anti-VEGF antibody Ranibizumab In the treatment of Neovascular Age-related macular

degeneration) clinical trials that established the superiority of ranibizumab as a treatment for nAMD reported their findings.7,8_{At 12 and 24 months,}_>_{90% of eyes treated with ranibizumab (0.5 mg) avoided} the loss of three lines of VA compared with<64% of eyes treated with PDT (the comparator in the ANCHOR trial) or 62% of eyes treated with sham injections (the comparator in the MARINA trial). Even more importantly, eyes treated with ranibizumab showed on average anincreasein acuity of between 5 and 10 letters read on an ETDRS vision chart with some 30% achieving 70 letters (Snellen equivalent of 6/12), a level of vision which is compatible with visually demanding tasks such as fluent reading and driving. These results exceeded all expectations, as trials of other therapeutic agents, including the VEGF inhibitor pegaptanib (Macugen, Pfizer),3,4_{had shown on average a}_reduction_{in acuity in treated eyes of} between two and three lines over 24 months.

Although the outcome after ranibizumab therapy was clearly impressive, this was achieved using an intensive dosing schedule of monthly injections of the drug into the vitreous cavity of the eye. The method of administration is an invasive procedure with the attendant risks of infection and iatrogenic eye trauma. In addition, the requirement for monthly attendance over many years poses serious challenges for elderly patients. Furthermore, such intensive treatment regimens also create difficulties in terms of resource implications for health service providers. A small study [PIER: a study of rhuFAB V2 (ranibizumab) in subjects with subfoveal choroidal neovascularization secondary to age-related macular degeneration], which used a less intensive dosing schedule of three 4-weekly injections of ranibizumab 0.5 mg followed by retreatment at fixed 3-monthly intervals, did not yield equivalent VA results as those observed in the MARINA and ANCHOR trials.9_{Although mean VA improved in the PIER study in the first 3 months in a} manner similar to that seen in ANCHOR and MARINA, it gradually decreased thereafter, dropping back by 12 months to the mean observed at baseline. By contrast, PrONTO [Prospective Optical Coherence Tomography (OCT) Imaging of Patients with Neovascular Age-Related Macular Degeneration Treated with intraOcular Ranibizumab], another small clinical trial of some 40 patients, suggested that a reduction in treatment frequency could be achieved through rigorous tailoring of treatment to morphological

(32)

of dosing interval would require continuous monitoring and tailoring of therapy. The study design of the ANCHOR and MARINA and other smaller randomised controlled trials (RCTs) that investigated effectiveness of ranibizumab also did not permit conclusions to be drawn about the total duration of treatment required.

Although the role of ranibizumab in the management of nAMD was being investigated, another drug [bevacizumab (Avastin®_{, Roche)] was identified by a number of small uncontrolled case series as having} equivalent visual benefits.11_{Remarkable improvements in acuity and morphological findings following} intravitreal injection of bevacizumab were reported by investigators from countries across the world,12,13 primarily because it was available and the cost was more affordable. Bevacizumab (a pan-VEGF

monoclonal antibody, as is pegaptanib) is the parent molecule from which ranibizumab was derived and the same manufacturer holds the patents and licences for both drugs. Bevacizumab was licensed for use in colorectal cancer, and the therapeutic dose for systemic administration for this condition is approximately 1000 times greater than that required for intraocular use. Thus clinicians were able to offer a cheap alternative to ranibizumab through unlicensed use of bevacizumab. The IVAN (Inhibit VEGF in Age-related choroidal Neovascularisation) trial was developed to test whether or not visual results obtained with bevacizumab were as good as those obtained with ranibizumab.

By virtue of their ability to inhibit all classes of VEGF, both ranibizumab and bevacizumab have the potential to induce serious ocular and systemic side effects. VEGF is known to have an important growth-promoting role in the retina14_{and is also thought to maintain the fenestrated phenotype of the} choroidal vasculature.15_{Therefore, there is concern that pan inhibition of VEGF over long periods of time} can cause atrophic changes in neural, retinal pigment epithelial (RPE) and vascular cells and tissues, with serious consequences for visual function.

Although no such adverse effects have been detected in clinical trials using VEGF inhibition

strategies,4,7–10,16the data available from trials relate to relatively short periods of follow-up, with few patients having been followed beyond 2 years. Repeated intraocular penetration for drug delivery carries a risk of endophthalmitis, traumatic cataract or retinal detachment. However, as shown by the VISION (pegaptanib for neovascular age-related macular degeneration) clinical trials, adherence to protocols that emphasise sterility and administration of the drug by experienced personnel reduces these risks to acceptable levels.4

Systemic side effects remain a concern as the pooled findings from ANCHOR and MARINA trials revealed a slight excess of thromboembolic events in the highest dose of ranibizumab groups and a small increase in non-ocular haemorrhages in the treatment groups.7–9In addition, circulating antibodies to ranibizumab were discernible in serum samples in a significant proportion of patients who received ocular administration of the drug.7,8_{As bevacizumab had not been tested in a controlled trial environment, there was no orderly} collection of information on its potential to cause systemic drug toxicity. Therefore, the IVAN trial also proposed the collection of serum samples from participants immediately prior to the Trial, and at the first post-injection visit, for the assay of VEGF, the levels of VEGF inhibitors themselves and to detect circulating antibodies to the inhibitors.

The IVAN trial also offered the opportunity for the creation of an accompanying biobank of serum and deoxyribonucleic acid (DNA). The recent advances in pharmacogenomics have revealed that genetic variation modifies the therapeutic response to drugs in a number of disease conditions.17_{Thus, there} is increasing enthusiasm for linking DNA biobanks to RCTs in which participants are extremely well

phenotyped. Trials using VEGF inhibition strategies in cancer have shown that both survival and toxicity are influenced by genetic variation.18_{It is plausible that similar mechanisms may influence visual outcomes} following therapeutic VEGF inhibition in choroidal neovascularisation (CNV). The IVAN trial therefore proposed the construction of a DNA biobank to test pharmacogenetic associations between treatment responsiveness and key genetic polymorphisms.

(33)

Summary of existing evidence

There was no systematic review of VEGF inhibitors in the treatment of nAMD due to AMD in 2007 when the IVAN trial was conceived. Neither was there a head-to-head comparison of the two main inhibitors of VEGF that were available at that time. There were no data on the minimum treatment frequency/duration that is required to maintain the maximal visual benefit achieved with either of the drugs studied in IVAN, and no trial had compared continuing monthly treatment with early cessation of VEGF inhibition with treatment being restarted if signs of lesion reactivation were detected. Therefore, the research questions that the IVAN study set out to address had not been investigated directly previously.

Importance of the health problem to the NHS

Epidemiological studies have shown that there are some 25,000 incident cases of nAMD each year in the UK.1_{RCTs had demonstrated the substantial benefit of ranibizumab for nAMD. Bevacizumab is considerably} cheaper than ranibizumab. The drug costs alone for monthly administration of ranibizumab were estimated to be about £11,000 per patient per year and the cost of assessments and treatment delivery about £2500 per year, with a potential annual cost to the UK NHS of up to £300M per year. The cost-effectiveness of VEGF inhibitor treatment is influenced greatly by the difference in drug cost, with ranibizumab being 20 times the price of bevacizumab. The costs of administering treatment are also high and there were no recommendations about the likely duration of treatment required.

The absence of robust information about the safety of bevacizumab, and uncertainty about treatment frequency for either drug, formed the basis for the alternative treatments in the IVAN trial.4_Accumulating evidence that susceptibility to nAMD is influenced by the carriage of specific polymorphisms in a number of genes which encode proteins involved in immune mediation and regulation,19–21and the fact that antibodies to the VEGF inhibitors had been found to develop over time, provided a strong rationale for the establishment of a DNA and serum biobanks in the IVAN trial. While the IVAN trial was being designed, The Comparison of Age-related macular degeneration Treatments Trials (CATT)22,23_{was developed in parallel in the USA. With the} acquiescence of the funding organisations of the two trials, agreement was reached to share information about the design and conduct of the trials, such as protocols and methods for the collection of adverse events (AEs). This exchange of information allowed CATT22_{and IVAN investigators to design the trials to facilitate} future meta-analyses of the outcomes of the two trials.

Aims and objectives

The aim of the IVAN trial was to investigate alternative VEGF inhibition treatment regimens for the treatment of nAMD. We hypothesised that:

l Bevacizumab is not inferior to ranibizumab with respect to the benefits of VEGF inhibition in maintaining/improving VA in eyes with nAMD.

l Treatment with VEGF inhibition can be‘safely’withdrawn at 3 months with monthly review to detect reactivation, that is, criteria for restarting treatment can be prespecified to prevent any difference in average VA compared with continuing monthly treatment.

(34)

The trial had three specific inter-related objectives:

l To estimate the:

¢ relative effectiveness of two VEGF inhibitors, namely ranibizumab and bevacizumab, on visual outcome in patients with nAMD; existing evidence of the benefit of VEGF inhibitors compared with sham treatment precludes inclusion of a sham VEGF inhibition arm

¢ effectiveness of more frequent compared with less frequent VEGF inhibition in improving or maintaining visual function, with stringent criteria for restarting treatment to prevent VA loss in patients receiving less frequent treatment

¢ cost-effectiveness of the alternative treatment strategies outlined above.

In addition to estimating the effectiveness of the treatments on visual outcome, the trial was designed to estimate differences in lesion morphology, quality of life and safety as secondary outcomes.

(35)

Chapter 2 Methods

Study design

The IVAN study was a multicentre, randomised controlled factorial trial. The trial is registered as ISRCTN92166560. The research objectives were addressed by randomising participants to one of four combinations of two treatment factors (Table 1). Participants, clinicians and trial personnel were masked to the VEGF inhibitor to which a participant was assigned. Pharmacies dispensed the appropriate drug to the ophthalmic clinic, as a pre-filled syringe for bevacizumab or in the commercially available vial for ranibizumab. We aimed to achieve masking of the drug by using unmasked study teams (ophthalmologists and nurses) who had no role in the outcome assessments in the trial. This method for masking had been used extensively in trials that involve invasive methods of treatment delivery, as it is ethically unjustified to inject a placebo into the eye because of the potential for serious adverse events (SAEs) related to the procedure itself. All assessments and treatment management decisions were made by other trial personnel who were masked to allocation to drug throughout the trial, and to allocation to treatment frequency until after the third treatment. We chose not to mask participants, clinicians and trial personnel to whether or not patients were allocated to continuous (i.e. monthly treatment) or discontinuous treatment (stop at 3 months with reinitiation of therapy if disease reactivation occurred).

Changes to study design after commencement of the study

The only change to the study design was a change to the method of masking the study drug, made in October 2007, before recruitment to the trial began. Bevacizumab was supplied in a pre-filled syringe and ranibizumab was supplied in a commercially manufactured vial. We originally intended that unmasked injectors, with no other role in the trial, would ensure masking. However, nine of the smaller sites did not have enough ophthalmologists to be able to guarantee the availability of an unmasked injector at every visit. For these sites, an unmasked ophthalmic nurse was allowed to draw up ranibizumab into a syringe identical to the pre-filled syringe of bevacizumab. Thus the injecting ophthalmologist was kept masked to the identity of the drug and could therefore be allowed to perform other study procedures. All other trial personnel involved in treatment decisions and outcome assessments remained masked. In December 2007, we further clarified that the injector must be an ophthalmologist; it was always the case that the injections should be performed by an ophthalmologist but we were asked to make this explicit in the protocol by the IVAN Data Monitoring and Safety Committee (DMSC).

TABLE 1 Factorial study design

Treatment regimen Ranibizumab Bevacizumab

Continue treatment @ 3 months A B

(36)

Participants

Eligibility criteria Inclusion criteria:

l adults of either sex, aged≥50 years

l newly referred for the treatment of nAMD in the first or second eye

l best corrected distance visual acuity (BCVA) of≥25 letters measured on a standard ETDRS chart l any component of the neovascular lesion (CNV, blood, serous pigment epithelial detachment, elevated

blocked fluorescence) involving the centre of the fovea.

Exclusion criteria:

l previous treatment with PDT or a VEGF inhibitor in the eye being considered for inclusion l argon laser treatment to the proposed study eye within the last 6 months

l long-standing CNV evidenced by the presence of fibrosis in excess of 50% of the total lesion l greatest linear diameter of>6000μm (equivalent to about 12 disc diameters)

l presence of thick blood involving the centre of the fovea

l presence of other active ocular disease causing concurrent vision loss, for example diabetic retinopathy l patients with eight or more dioptres of myopia

l pregnant and/or lactating women

l women with childbearing potential (i.e. not sterilised or not post-menopausal) who are unwilling to use contraception

l men with a spouse or partner with childbearing potential unless the participant has agreed to use condoms.

A past medical history of cardiovascular disease (CVD) or cardiovascular comorbidity, for example previous myocardial infarction (MI), stroke or current angina, was not an exclusion criterion. However, such conditions were documented carefully at the time of recruitment, and the potential benefits and harms of treatment were discussed carefully with potential participants.

The eligibility criteria included some items that related to the proposed study eye and some items that related to the person being screened for eligibility. The trial studied only one eye of each participant (eligible‘study eye’in an eligible person). If a fellow eye developed nAMD during the trial, it was treated with the optimum locally available treatment.

Changes to study eligibility criteria after commencement of the study

‘Patients with 8 or more dioptres of myopia’was added as an exclusion criterion in June 2008 to ensure consistency with other trials, such as the ANCHOR and MARINA trials.7,8

Settings

Patients were recruited to the IVAN trial from 23 ophthalmic units in NHS hospital trusts.